Obstructive sleep apnea (OSA) occurs in 50-100% of patients with Down syndrome and can significantly cause and exacerbate medical problems in these patients. Current surgical management in these children is imperfect. There are variable surgical success rates for both first line surgery [palatine tonsillectomy and adenoidectomy (T&A)] as well as secondary surgeries performed if and when T&A fails. There is a critical need for a diagnostic modality that takes into account airway anatomy, tissue compliance, and collapsibility to be able to predict surgical outcome and improve surgical planning in these patients. Our central hypothesis is that upper airway flow-structure interaction (FSI) modeling using three-dimensional (3-D) computational simulations from dynamic MRI data can be used to predict surgical outcome for children with Down syndrome who have OSA despite previous T&A. The long-term goal is to improve surgical outcome of children with Down syndrome and OSA by creating an accurate FSI predictive model. Such a diagnostic tool would help tailor surgical procedures to be more effective as well as identify and avoid unnecessary (or unhelpful) surgical procedures. These created models then can also be adjusted and applied to other populations with OSA. Our specific aims include: 1) In children with Down syndrome with persistent OSA despite previous T&A, collect data characterizing upper airway anatomy, tissue compliance, and collapsibility; 2) Generate and validate individualized dynamic FSI models for each child in specific aim 1; and 3) Using the validated dynamic computational models, predict the success of surgical treatment on children with Down syndrome who have persistent OSA despite previous T&A. This work is innovative as it uses dynamic rather than static MR imaging data and applies a unique computational model that accurately captures the unsteadiness of the flow and accounts for the interaction between the airflow and the surrounding airway flexible structures.